Example:

My November bill was easy to figure since the average temps. were the same for each year.

2006
83 therms, 29 days in billing cycle, average temp was 38 F
2007
58 therms, 30 days in billing cycle, average temp was 38 F

I clearly used less gas in Nov 07 compared to Nov 06.

My December bill is harder to compare to last year since it was much colder this year. Again, what's tripping me up is how you take into account the difference in average temp for the two months being compared.

2006
119 therms, 31 days in billing cycle, average temp was 28 F
2007
135 therms, 30 days in billing cycle, average temp was 16 F

Do you see what I'm trying to do? Is there a way to figure this out?

Kevin

Using your numbers for December 2006:

119 therms / 31 days = 3.8 therms per day

Now to make this relevant to the average temperature that month, I would divide this derived figure by the average temperature to get an average therms per day per degree average temperature.

3.8 therms per day / 28F = 0.14 therms per day per degree F

Doing the same calculations for December 2007 gives me the figure 0.28 therms per day per degree F.

This method should give you a rough comparison, but it's not ideal in some ways. I don't like using averages, they can be heavily skewed over wide ranges. Also, what we are comparing is your gas consumption against the average temperature. Your heating does not however heat that way, it raises the temperature in your house to whatever the thermostat is set at above that temperature, which means we're missing a more useful number from our calculations.

For example, when it was on average 16F, and your thermostat is set at 70F, your heating had to raise the temperature in the house on average by 54F. This number has a direct relationship to how much work your central heating did, and so would be a better way to compare the numbers.

If I used the assumed figure of 70F for the thermostat, I get the numbers:

December 2006: 0.09 therms per degree F your house temperature is raised.

December 2007: 0.08 therms per degree F your house temperature is raised.

If however your thermostat was reduced to 65F during December 2007, the figure would become 0.09 therms per degree your house temperature is raised.

As you can see, the amount of work performed by your central heating is very important to these figures and could turn them around entirely. Unfortunately, if your house is anything like ours, the thermostat setting changes several times a day. Traditional mercury thermostats are also exceptionally inaccurate. Some by 5F or more, which is perhaps why some people report that they only heat their house to 64F!

We might also have to factor in other appliances that use gas in your house?

To really be sure if there is a statistical difference we would have to collect all the data you have for your gas usage, the weather and any other data you have and perform some type of statistical test on it, which would give us a probability of the difference in gas consumption being by chance. It's kind of extreme, I certainly haven't bothered to statistical test any of my own bills yet, but it would give you as close to an accurate answer as possible.

Anyway, I hope this highlights the difficulty we have deriving useful figures for energy use outside controlled environments. As the variables increase and the accuracy of the figures we do have are questioned, attempts to turn them in to something useful become more difficult.

Kevin

Maybe you misunderstood. The 28 F temp is outside, not inside. That would be a bit chilly even for me.

Thanks for the invite, but I can't imagine a worse place to live for me. We might have some cold winters, but you have that heat and humidity for a good hunk of the year. Nothing against FL, but I love the seasons and the huge swings in temps. It makes life interesting.

Right now it's a balmy 27 F, but by tonight we should have temps in the -15 F range with wind chills in the -40 F. range. A good night to watch a movie, eat something good and look through the garden catalogs. Winter can be very cozy.

K